# Compensation philosophy file template The compensation-benchmark skill calibrates survey data against the firm's compensation philosophy. This file is the philosophy. Copy the JSON below to `philosophy.json` (or wherever your skill config points), fill it in, and version it in git. The philosophy is rarely revised — usually annually at most, often less. When it changes, every benchmark recommendation post-change uses the new philosophy. The skill captures the philosophy version in the audit record so future audits can reproduce the recommendation. ## JSON shape ```json { "philosophy_version": "2026.1", "effective_from": "2026-01-10", "approver": "Comp Committee minutes 2026-01-08", "target_percentiles": { "engineering": { "base": 60, "equity": 75, "bonus_or_ote": 50 }, "sales": { "base": 50, "equity": 50, "bonus_or_ote": 75 }, "go_to_market_other": { "base": 60, "equity": 60, "bonus_or_ote": 60 }, "g_and_a": { "base": 60, "equity": 50, "bonus_or_ote": 50 } }, "band_widths_pct": { "default": 10, "by_level": { "junior": 8, "senior": 12, "staff": 15, "principal": 18, "executive": 25 } }, "level_mapping": { "firm_to_radford": { "L3": "P3", "L4": "P4", "L5": "P4", "L6": "P5", "L7": "P6", "L8": "P7" }, "firm_to_pave": { "L3": "Mid IC", "L4": "Senior IC", "L5": "Senior IC", "L6": "Staff IC", "L7": "Principal IC", "L8": "Senior Principal IC" } }, "geography_adjustments": { "remote_us": 0, "remote_intl": -15, "san_francisco_msa": 0, "new_york_msa": 0, "seattle_msa": -3, "austin_msa": -10, "london": -8, "toronto": -12, "remote_latam": -35 }, "current_strike_price_usd": 5.20, "vesting_period_years": 4, "equity_grant_type": "RSU", "exception_band": { "max_above_top": 15, "approval_required": "comp_committee" }, "public_range_policy": { "minimum_band_width_pct": 15, "include_equity_target_dollar": true, "include_bonus_target_dollar": true }, "pay_equity_audit_cadence_months": 12, "last_pay_equity_audit": "2026-02-15" } ``` ## Field-by-field ### `philosophy_version`, `effective_from`, `approver` Versioning. The skill captures `philosophy_version` in the audit record so the recommendation is reproducible against a specific version of this file. `effective_from` is the date the philosophy applies to NEW recommendations — recommendations made before that date used the prior philosophy. `approver` cites the approval source (Comp Committee minutes, board resolution, etc.). ### `target_percentiles` Per function family, the target percentile per component. The most common patterns: - **Engineering**: 60th base, 75th equity (founder-friendly equity to attract IC talent), 50th bonus. - **Sales**: 50th base, 50th equity, 75th OTE (the variable comp is the lever). - **G&A**: 60th base across components (predictable, market-rate). If the firm's strategy is "we pay top of market across the board," set everything to 75th. If the firm's strategy is "we pay base at market and over-index on equity," set base to 50th and equity to 75th-90th. ### `band_widths_pct` The recommended band width as a percentage of the midpoint. Default 10% (recommended midpoint ±10%). Per-level overrides absorb the wider individual variance at senior levels. If a single number is too rigid, the skill respects the per-level overrides. ### `level_mapping` Mapping from the firm's internal ladder to each survey's ladder. The skill cannot infer this — it has to be specified per survey the firm uses. If a level mapping is missing, the skill halts and asks the user to add it. This is the single most-edited part of the philosophy; it's also the most consequential, because the wrong mapping shifts every recommendation by a percentile-band. ### `geography_adjustments` Per-geography multiplier. `0` means use the survey's value for that geography directly. `-15` means apply a 15% reduction (e.g. for `remote_intl`). The adjustments must be defensible — random adjustments here are how pay-equity gaps creep in. If the firm has a published location-based pay policy, this section should match it line-for-line. ### `current_strike_price_usd`, `vesting_period_years`, `equity_grant_type` Used for converting annualized survey equity values to grant size. Strike price is the most-recent 409A or option-grant strike. Vesting is typically 4 years. Grant type matters for tax framing but not for the band math. ### `exception_band` When the skill is asked about a band-exception ("can we offer above the top?"), the philosophy says how high (`max_above_top: 15` means up to 15% above the top of the recommended band) and who approves. The skill itself does NOT approve exceptions; it surfaces the policy. ### `public_range_policy` Compliance posture for NYC LL 32-A, CO/CA/WA pay-transparency requirements. `minimum_band_width_pct: 15` is the firm's "good faith" floor — the skill warns if a recommended band falls below this width. ### `pay_equity_audit_cadence_months`, `last_pay_equity_audit` For the audit-record metadata. The skill notes the cadence and last audit so the recommendation can be flagged if the firm is overdue for an equity audit. ## When to revise the philosophy - **Strategy shift** — the firm decides to over-index on equity vs. cash. Update target percentiles. - **New geography** — opening a new region. Add to `geography_adjustments` based on local market data. - **New survey added** — add a level mapping for the new survey. - **Pay-equity audit findings** — if the audit surfaces gaps, the philosophy may need revision (band widths, geography adjustments). Each revision bumps `philosophy_version`. Old audit records remain interpretable against their respective version. ## What the philosophy is NOT - It is NOT a candidate-by-candidate negotiation guide. - It is NOT a one-time setup; it evolves with the firm. - It is NOT confidential to the recruiter — the philosophy should be visible to every hiring manager, ideally documented internally for transparency.